The basic immunoglobulin structural unit in vertebrate systems is composed of two identical "light" polypeptide chains of molecular weight approximately 23,000 daltons, and two identical "heavy" chains of molecular weight 53,000-70,000. The four chains are joined by disulfide bonds in a "Y" configuration in which the light chains bracket the heavy chains starting at the mouth of the Y and continuing through the divergent region or "branch" portion which is designated the Fab region. Heavy chains are classified as gamma (.gamma.), mu (.mu.), alpha (.alpha.), delta (.delta.), or epsilon (.epsilon.), with some subclasses among them which vary according to species; and the nature of this chain, which has a long constant region, determines the "class" of the antibody as IgG, IgM, IgA, IgD, or IgE, respectively. Light chains are classified as either kappa (.kappa.) or lambda (.lambda.). Each heavy chain class can be associated with either a kappa or lambda light chain. The light and heavy chains are covalently bonded to each other, and the "tail" portions of the two heavy chains are bonded to each other by covalent disulfide linkages when the immunoglobulins are generated either by hybridomas or by B cells.
The amino acid sequence of each immunoglobulin chain runs from the N-terminal end at the top of the Y to the C-terminal end at the bottom. The N-terminal end contains a variable region (V) which is specific for the antigen to which it binds and is approximately 100 amino acids in length, there being variations between light and heavy chain and from antibody to antibody. The variable region is linked in each chain to a constant region (C) which extends the remaining length of the chain. Linkage is seen, at the genomic level, as occurring through a linking sequence known as the joining (J) region in the light chain gene, which encodes about 12 amino acids, and as a combination of diversity (D) region and joining (J) region in the heavy chain gene, which together encode approximately 25 amino acids. The remaining portions of the chain, the constant regions, do not vary within a particular class with the specificity of the antibody (i.e., the antigen to which it binds). The constant region or class determines subsequent effector function of the antibody, including activation of complement and other cellular responses, while the variable region determines the antigen with which it will react.
Since the development of the cell fusion technique for the production of monoclonal antibodies by Kohler and Milstein, many individual immunoglobulin species have been produced in quantity. Most of these monoclonal antibodies are produced in a murine system and, therefore, have limited utility as human therapeutic agents unless modified in some way so that the murine monoclonal antibodies are not "recognized" as foreign epitopes and "neutralized" by the human immune system.
One approach to this problem has been to attempt to develop human or "humanized" monoclonal antibodies, which are "recognized" less well as foreign epitopes and may overcome the problems associated with the use of monoclonal antibodies in humans. Applications of human B cell hybridoma-produced monoclonal antibodies hold great promise for the treatment of cancer, viral and microbial infections, B cell immunodeficiencies with diminished antibody production, and other diseases and disorders of the immune system.
However, several obstacles exist with respect to the development of human monoclonal antibodies. For example, with respect to monoclonal antibodies which recognize human tumor antigens for the diagnosis and treatment of cancer, many of these tumor antigens are not recognized as foreign antigens by the human immune system and, therefore, these antigens may not be immunogenic in man.
Another problem with human monoclonal antibodies is that most such antibodies obtained in cell culture are of one class or isotype, the IgM type. Under certain circumstances, monoclonal antibodies of one isotype might be more preferable than those of another in terms of their diagnostic or therapeutic efficacy since, as noted above, the isotype determines subsequent effector function of the antibody, including activation of complement and other cellular responses. For example, from studies on antibody-mediated cytolysis it is known that unmodified mouse monoclonal antibodies of subtype .gamma.2a and .gamma.3 are generally more effective in lysing target cells than are antibodies of the .gamma.1 isotype. This differential efficacy is thought to be due to the ability of the .gamma.2a and .gamma.3 subtypes to more actively participate in the cytolytic destruction of the target cells. Particular isotypes of a murine monoclonal antibody can be prepared either directly, by selecting from the initial fusion, or secondarily, from a parental hybridoma secreting monoclonal antibody of a different isotype, by using the "sib selection" technique to isolate class-switch variants (Steplewski et al., 1985, Proc. Natl. Acad. Sci. USA 82:8653; Spira et al., 1984, J Immunological Methods 74:307.
When human monoclonal antibodies of the IgG type are desired, however, it has been necessary to use such tedious techniques as cell sorting, to identify and isolate the few cells which are producing antibodies of the IgG or other type from the majority producing antibodies of the IgM type. A need therefore exists for an efficient method of switching antibody classes in isolated antibody-producing cells for any given antibody of a predetermined or desired antigenic specificity.
Various solutions to these problems with monoclonal antibodies for human use have been developed based on recent methods for the introduction of DNA into mammalian cells to obtain expression of immunoglobulin genes, particularly for production of chimeric immunoglobulin molecules comprising a human and a non-human portion. More specifically, the antigen combining (variable) region of the chimeric antibody is derived from a non-human source (e.g., murine), and the constant region of the chimeric antibody (which confers biological effector function to the immunoglobulin) is derived from a human source. Such "humanized" chimeric antibodies should have the antigen binding specificity of the non-human antibody molecule and the effector function conferred by the human antibody molecule.
Generally, chimeric antibodies have been produced conventionally by procedures comprising the following steps (although not necessarily in this order): (1) identifying and cloning the gene segment encoding the antigen binding portion of the antibody molecule; this gene segment (VDJ for heavy chains or VJ for light chains, or more simply, the variable region) may be obtained from either a cDNA or genomic source; (2) cloning the gene segments encoding the constant region or desired part thereof; (3) ligating the variable region with the constant region so that the complete chimeric antibody is encoded in a transcribable and translatable form; (4) ligating this construct into a vector containing a selectable marker and appropriate gene control regions; (5) amplifying this construct in bacteria; (6) introducing the DNA into eukaryotic cells (by transfection), most often cultured mammalian cells such as lymphocytes; (7) selecting for cells expressing the selectable marker; (8) screening for cells expressing the desired chimeric antibody; and (9) testing the antibody for appropriate binding specificity and effector functions.
Antibodies of several distinct antigen binding specificities have been manipulated by these procedures to produce chimeric proteins. In addition several different effector functions have been achieved by linking new sequences to those encoding the antigen binding region. Some of these include enzymes (Neuberger et al., 1984, Nature 312:604), immunoglobulin constant regions from another species and constant regions of another immunoglobulin chain (Sharon et al., 1984, Nature 309:364; Tan et al., 1985, J. Immunol. 135:3565-3567). Neuberger et al., PCT Publication WO 86/01533 (1986) also discloses production of chimeric antibodies and mentions, among the technique's many uses, the concept of "class switching."
Cabilly, et al., U.S. Pat. No. 4,816,567 issued Mar. 28, 1989, describes altered and native immunoglobulins, including constant-variable region chimeras, which are prepared in recombinant cell culture. The immunoglobulins contain variable regions which are immunologically capable of binding predetermined antigens. The vectors and methods disclosed are suitable for use in various host cells including a wide range of prokaryotic and eukaryotic organisms.
Fell et al., U.S. Pat. No. 5,202,238 issued Apr. 13, 1993, describes a process for producing chimeric antibodies using recombinant DNA vectors and homologous recombination in vivo. The process uses novel recombinant DNA vectors to engineer targeted gene modification accomplished via homologous recombination in either (a) cell lines that produce antibodies having desired antigen specificities, so that the antigen combining site of an antibody molecule remains unchanged, but the constant region of the antibody molecule, or a portion thereof, is replaced or modified; or (b) cell lines that produce antibodies of desired classes which may demonstrate desired effector functions, so that the constant region of an antibody molecule remains unchanged, but the variable region of the antibody molecule or a portion thereof, is replaced or modified. The reported efficiency of recombination was relatively low, however, ranging from 0.39% to 0.75% in several attempts to replace a mouse heavy chain constant region with a human counterpart, even when a selectable marker gene was used to recover recombinant genomes.
Kucherlapati et al., in PCT Publication WO91/10741 (published Jul. 25, 1991) and in PCT Publication WO 94/02602 (published Feb. 3, 1994) disclose xenogeneic specific binding proteins or antibodies produced in a non-primate viable mammalian host by immunization of the mammalian host with an appropriate immunogen. In particular these publications disclose production of antigen-specific human monoclonal antibodies from mice engineered with loci for human immunoglobulin heavy and light chains using yeast artificial chromosomes (YACs). Such mice produce completely human antibodies in response to immunization with any antigen normally recognized by the mouse immune system, including human antigens; and B-cells from these mice are used to make hybridomas producing human monoclonal antibodies via conventional hybridoma production methods. While production of completely human antibodies from transgenic murine hybridomas solves many of the previous problems of producing human monoclonal antibodies, it may be desirable to modify the loci in the cells which produce the antibodies, for instance, to enhance expression or to alter an effector function.
Sauer et al, U.S. Pat. No. 4,959,317, issued Sep. 25, 1990, describes a method for producing site-specific recombination of DNA in eukaryotic cells at sequences designated lox sites. DNA sequences comprising first and second lox sites are introduced into eukaryotic cells and interacted with a recombinase designated Cre (typically, by transient expression of the recombinase from a plasmid), thereby producing recombination at the lox sites. Exemplified eukaryotic cells included yeast cells and monolayer cultures of a mouse cell line. Frequencies of Cre-mediated recombination ranged from 2-3% to 22% for repeated recombination attempts between an exemplary virus and a plasmid, and 98% for the case of deletion of a yeast leu2 gene flanked by lox sites. However, neither recombination in lymphoid cells nor manipulation of immunoglobulin loci is disclosed by Sauer et al.
Gu et al., 1993, Cell 73:1155-1164 describes a method to generate a mouse strain in which the J region and the intron enhancer in the heavy-chain locus are deleted from embryonic stem cells using Cre-mediated site-specific recombination. They then analyzed the immunoglobulin isotypes formed by recombination in heterozygous mutant B cells, activated with LPS plus IL4. The authors used Cre-mediated site-specific recombination merely to modify the heavy chain locus in stem cells which are not antibody-producing cells, and only to modify that locus by deletion, rather than to produce new combinations of gene sequences in antibody-producing cells to produce chimeric or modified antibodies.
Johnson et al., in PCT Publication WO93/19172 (published Sep. 9, 1993), and Waterhouse et al., in the journal article, Nucleic Acids Res (1993) 21:2265-2266, describe use of Cre-mediated site-specific recombination to effect recombinations for creation of a combinatorial library of antibodies in phage vectors. In the latter publication, Johnson, Waterhouse and coworkers describe two exemplary phage vectors designated A and B, where A encodes the light chain of a first antibody (and the heavy chain from a different antibody) and B encodes the heavy chain of the first antibody. In both vectors the variable heavy chain (VH) genes are flanked by two loxP sites, one of which is a mutant loxP site which prevents recombination within the vector from merely excising the VH genes. When Cre recombinase is provided in vivo by infecting the E. coli with a phage expressing Cre, vectors A and B can co-integrate by recombination between either mutant or wild-type loxP sites to create chimeric plasmids. Further recombination can then occur between the two wild-type or the two mutant loxP sites, to generate original vectors A and B or two new vectors, E and F. The heavy chains of A and B are therefore exchanged in E and F, and E encodes the Fab fragment of the first antibody for display as a fusion to the N-terminus of the phage gene 3 protein (g3p). The authors indicate that the method should allow the creation of extremely large combinatorial repertoires of phage-expressing antibodies, for example by providing a light chain repertoire in phage vector A and a heavy chain repertoire in phage vector B. However, these two publications do not disclose use of Cre-mediated site-specific recombination for any other purpose besides exchange of heavy chain genes in the particular bacteriophage vectors disclosed and, more specifically, do not propose any modification of immunoglobulin sequences in the genome of any antibody-producing cell.
Accordingly, the present invention provides for a novel use of Cre-mediated site-specific recombination for modification of immunoglobulin loci in the cellular genome of any antibody-producing cell.